| Message ID | 20250218195048.74692-1-kuba@kernel.org (mailing list archive) |
|---|---|
| Headers | show |
| Series | selftests: drv-net: improve the queue test for XSK | expand |
On 02/18, Jakub Kicinski wrote: > We see some flakes in the the XSK test: > > Exception| Traceback (most recent call last): > Exception| File "/home/virtme/testing-18/tools/testing/selftests/net/lib/py/ksft.py", line 218, in ksft_run > Exception| case(*args) > Exception| File "/home/virtme/testing-18/tools/testing/selftests/drivers/net/./queues.py", line 53, in check_xdp > Exception| ksft_eq(q['xsk'], {}) > Exception| KeyError: 'xsk' > > I think it's because the method or running the helper in the background > is racy. Add more solid infra for waiting for a background helper to be > initialized. > > Jakub Kicinski (4): > selftests: drv-net: use cfg.rpath() in netlink xsk attr test > selftests: drv-net: add a way to wait for a local process > selftests: drv-net: improve the use of ksft helpers in XSK queue test > selftests: drv-net: rename queues check_xdp to check_xsk > > .../selftests/drivers/net/xdp_helper.c | 22 ++++++- > tools/testing/selftests/drivers/net/queues.py | 55 ++++++++---------- > tools/testing/selftests/net/lib/py/ksft.py | 5 ++ > tools/testing/selftests/net/lib/py/utils.py | 58 +++++++++++++++++-- > 4 files changed, 103 insertions(+), 37 deletions(-) > > -- > 2.48.1 > Acked-by: Stanislav Fomichev <sdf@fomichev.me>
On Tue, Feb 18, 2025 at 11:50:44AM -0800, Jakub Kicinski wrote: > We see some flakes in the the XSK test: > > Exception| Traceback (most recent call last): > Exception| File "/home/virtme/testing-18/tools/testing/selftests/net/lib/py/ksft.py", line 218, in ksft_run > Exception| case(*args) > Exception| File "/home/virtme/testing-18/tools/testing/selftests/drivers/net/./queues.py", line 53, in check_xdp > Exception| ksft_eq(q['xsk'], {}) > Exception| KeyError: 'xsk' > > I think it's because the method or running the helper in the background > is racy. Add more solid infra for waiting for a background helper to be > initialized. Sorry for the raciness I introduced. I think overall the direction this is going is good - thanks for cleaning this up. Please see my comments on patch 2 and the hang introduced with this change.
We see some flakes in the the XSK test: Exception| Traceback (most recent call last): Exception| File "/home/virtme/testing-18/tools/testing/selftests/net/lib/py/ksft.py", line 218, in ksft_run Exception| case(*args) Exception| File "/home/virtme/testing-18/tools/testing/selftests/drivers/net/./queues.py", line 53, in check_xdp Exception| ksft_eq(q['xsk'], {}) Exception| KeyError: 'xsk' I think it's because the method or running the helper in the background is racy. Add more solid infra for waiting for a background helper to be initialized. Jakub Kicinski (4): selftests: drv-net: use cfg.rpath() in netlink xsk attr test selftests: drv-net: add a way to wait for a local process selftests: drv-net: improve the use of ksft helpers in XSK queue test selftests: drv-net: rename queues check_xdp to check_xsk .../selftests/drivers/net/xdp_helper.c | 22 ++++++- tools/testing/selftests/drivers/net/queues.py | 55 ++++++++---------- tools/testing/selftests/net/lib/py/ksft.py | 5 ++ tools/testing/selftests/net/lib/py/utils.py | 58 +++++++++++++++++-- 4 files changed, 103 insertions(+), 37 deletions(-)